ClawHub
Back to skill

Security audit

Locke

Security checks across malware telemetry and agentic risk

Overview

This is a low-risk John Locke roleplay skill with no tools, code execution, credentials, or persistence, though its broad name triggers can cause accidental activation.

Safe to install for educational roleplay. Use explicit invocation such as /locke when possible, and treat responses as an AI simulation of John Locke rather than a neutral assistant or real historical source.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
90% confidence
Finding
The trigger phrase "locke" is generic enough to match ordinary discussion about the philosopher, a surname, or unrelated content, which can cause unintended skill invocation. In this skill, the impact is limited because model invocation is disabled and no tools are available, but accidental activation can still hijack routing, confuse users, or suppress the intended default assistant behavior.

Overly Broad Trigger

Low
Category
Trigger Abuse
Confidence
83% confidence
Finding
The trigger "洛克" is very short and likely to appear in normal Chinese text referring to John Locke, names, or educational discussion, making accidental activation plausible. Context reduces severity because the skill has no tools and cannot invoke a model, but unintended routing or persona takeover remains a real integrity and usability issue.

VirusTotal

63/63 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.